Video data labeling
The task itself requires to mark the OCR'ed text found in the image and annotate with the language the text is in. Separate queues will be established for each language that has been automatically detected. While most of the text in the images is expected to be in the target language, this task involves human review following the automated recognition.